PURPOSE: Estimating the value of immuno-oncology (IO) treatments is challenging. It is common for the OS benefit to represent a major source of uncertainty due to the immaturity of data from clinical trials. Typically extrapolation of OS relies on fitting a parametric curve under a set of standard distribution assumptions. The emergence of IO agents and their new mechanisms of action add further layers of complexity to OS extrapolation. The objective of this workshop is to discuss additional challenges in fitting survival models for IO treatments and evaluating emerging approaches to overcome existing challenges from both statistical and clinical perspectives. The workshop will build upon previous workshops held at 2016 and 2017 ISPOR meetings, assessing survival projection methods for IO agents. DESCRIPTION: Both standard and emerging methodological approaches used to date will be applied to the same dataset from a clinical trial of an IO agent to fit the observed OS data and to extrapolate the OS outcome beyond the trial. The model performance and extrapolations will be compared and contrasted. Based on the empirical findings, IO-specific data gaps and uncertainties, alongside strengths and weaknesses, will be examined and recommendations will be provided for additional methodological considerations.